Merge commit 'mem'

[public-inbox.git] / lib / PublicInbox / WWW.pm

diff --git a/lib/PublicInbox/WWW.pm b/lib/PublicInbox/WWW.pm
index a2c2a4a6ff2944359ad2a5012c8f2f915279183e..c1c392657b96dbe023e9a70a1547c3092cf8bacb 100644 (file)
--- a/lib/PublicInbox/WWW.pm
+++ b/lib/PublicInbox/WWW.pm
@@ -36,6 +36,17 @@ sub run {
        PublicInbox::WWW->new->call($req->env);
 }
 
+my %path_re_cache;
+
+sub path_re ($) {
+       my $sn = $_[0]->{SCRIPT_NAME};
+       $path_re_cache{$sn} ||= do {
+               $sn = '/'.$sn unless index($sn, '/') == 0;
+               $sn =~ s!/\z!!;
+               qr!\A(?:https?://[^/]+)?\Q$sn\E(/[^\?\#]+)!;
+       };
+}
+
 sub call {
        my ($self, $env) = @_;
        my $ctx = { env => $env, www => $self };
@@ -50,14 +61,15 @@ sub call {
        } split(/[&;]+/, $env->{QUERY_STRING});
        $ctx->{qp} = \%qp;
 
-       my $path_info = $env->{PATH_INFO};
+       # not using $env->{PATH_INFO} here since that's already decoded
+       my ($path_info) = ($env->{REQUEST_URI} =~ path_re($env));
        my $method = $env->{REQUEST_METHOD};
 
        if ($method eq 'POST') {
-               if ($path_info =~ m!$INBOX_RE/(git-upload-pack)\z!) {
-                       my $path = $2;
+               if ($path_info =~ m!$INBOX_RE/(?:(\d+)/)?(git-upload-pack)\z!) {
+                       my ($part, $path) = ($2, $3);
                        return invalid_inbox($ctx, $1) ||
-                               serve_git($ctx, $path);
+                               serve_git($ctx, $part, $path);
                } elsif ($path_info =~ m!$INBOX_RE/!o) {
                        return invalid_inbox($ctx, $1) || mbox_results($ctx);
                }
@@ -77,10 +89,10 @@ sub call {
                invalid_inbox($ctx, $1) || get_atom($ctx);
        } elsif ($path_info =~ m!$INBOX_RE/new\.html\z!o) {
                invalid_inbox($ctx, $1) || get_new($ctx);
-       } elsif ($path_info =~ m!$INBOX_RE/
+       } elsif ($path_info =~ m!$INBOX_RE/(?:(\d+)/)?
                                ($PublicInbox::GitHTTPBackend::ANY)\z!ox) {
-               my $path = $2;
-               invalid_inbox($ctx, $1) || serve_git($ctx, $path);
+               my ($part, $path) = ($2, $3);
+               invalid_inbox($ctx, $1) || serve_git($ctx, $part, $path);
        } elsif ($path_info =~ m!$INBOX_RE/([\w-]+).mbox\.gz\z!o) {
                serve_mbox_range($ctx, $1, $2);
        } elsif ($path_info =~ m!$INBOX_RE/$MID_RE/$END_RE\z!o) {
@@ -91,13 +103,13 @@ sub call {
                invalid_inbox_mid($ctx, $1, $2) || get_attach($ctx, $idx, $fn);
        # in case people leave off the trailing slash:
        } elsif ($path_info =~ m!$INBOX_RE/$MID_RE/(T|t)\z!o) {
-               my ($inbox, $mid, $suffix) = ($1, $2, $3);
+               my ($inbox, $mid_ue, $suffix) = ($1, $2, $3);
                $suffix .= $suffix =~ /\A[tT]\z/ ? '/#u' : '/';
-               r301($ctx, $inbox, $mid, $suffix);
+               r301($ctx, $inbox, $mid_ue, $suffix);
 
        } elsif ($path_info =~ m!$INBOX_RE/$MID_RE/R/?\z!o) {
-               my ($inbox, $mid) = ($1, $2);
-               r301($ctx, $inbox, $mid, '#R');
+               my ($inbox, $mid_ue) = ($1, $2);
+               r301($ctx, $inbox, $mid_ue, '#R');
 
        } elsif ($path_info =~ m!$INBOX_RE/$MID_RE/f/?\z!o) {
                r301($ctx, $1, $2);
@@ -164,19 +176,20 @@ sub invalid_inbox ($$) {
 
 # returns undef if valid, array ref response if invalid
 sub invalid_inbox_mid {
-       my ($ctx, $inbox, $mid) = @_;
+       my ($ctx, $inbox, $mid_ue) = @_;
        my $ret = invalid_inbox($ctx, $inbox);
        return $ret if $ret;
 
-       $ctx->{mid} = $mid;
-       if ($mid =~ /\A[a-f0-9]{40}\z/) {
-               # this is horiffically wasteful for legacy URLs:
-               if ($mid = mid2blob($ctx)) {
-                       require Email::Simple;
-                       use PublicInbox::MID qw/mid_clean/;
-                       my $s = Email::Simple->new($mid);
-                       $ctx->{mid} = mid_clean($s->header('Message-ID'));
-               }
+       my $mid = $ctx->{mid} = uri_unescape($mid_ue);
+       my $ibx = $ctx->{-inbox};
+       if ($mid =~ m!\A([a-f0-9]{2})([a-f0-9]{38})\z!) {
+               my ($x2, $x38) = ($1, $2);
+               # this is horrifically wasteful for legacy URLs:
+               my $str = $ctx->{-inbox}->msg_by_path("$x2/$x38") or return;
+               require Email::Simple;
+               my $s = Email::Simple->new($str);
+               $mid = PublicInbox::MID::mid_clean($s->header('Message-ID'));
+               return r301($ctx, $inbox, mid_escape($mid));
        }
        undef;
 }
@@ -208,12 +221,6 @@ sub get_index {
        }
 }
 
-# just returns a string ref for the blob in the current ctx
-sub mid2blob {
-       my ($ctx) = @_;
-       $ctx->{-inbox}->msg_by_mid($ctx->{mid});
-}
-
 # /$INBOX/$MESSAGE_ID/raw                    -> raw mbox
 sub get_mid_txt {
        my ($ctx) = @_;
@@ -357,7 +364,7 @@ sub legacy_redirects {
 }
 
 sub r301 {
-       my ($ctx, $inbox, $mid, $suffix) = @_;
+       my ($ctx, $inbox, $mid_ue, $suffix) = @_;
        my $obj = $ctx->{-inbox};
        unless ($obj) {
                my $r404 = invalid_inbox($ctx, $inbox);
@@ -366,7 +373,11 @@ sub r301 {
        }
        my $url = $obj->base_url($ctx->{env});
        my $qs = $ctx->{env}->{QUERY_STRING};
-       $url .= (mid_escape($mid) . '/') if (defined $mid);
+       if (defined $mid_ue) {
+               # common, and much nicer as '@' than '%40':
+               $mid_ue =~ s/%40/@/g;
+               $url .= $mid_ue . '/';
+       }
        $url .= $suffix if (defined $suffix);
        $url .= "?$qs" if $qs ne '';
 
@@ -376,9 +387,9 @@ sub r301 {
 }
 
 sub msg_page {
-       my ($ctx, $inbox, $mid, $e) = @_;
+       my ($ctx, $inbox, $mid_ue, $e) = @_;
        my $ret;
-       $ret = invalid_inbox_mid($ctx, $inbox, $mid) and return $ret;
+       $ret = invalid_inbox_mid($ctx, $inbox, $mid_ue) and return $ret;
        '' eq $e and return get_mid_html($ctx);
        'T/' eq $e and return get_thread($ctx, 1);
        't/' eq $e and return get_thread($ctx);
@@ -393,8 +404,11 @@ sub msg_page {
 }
 
 sub serve_git {
-       my ($ctx, $path) = @_;
-       PublicInbox::GitHTTPBackend::serve($ctx->{env}, $ctx->{git}, $path);
+       my ($ctx, $part, $path) = @_;
+       my $env = $ctx->{env};
+       my $ibx = $ctx->{-inbox};
+       my $git = defined $part ? $ibx->git_part($part) : $ibx->git;
+       $git ? PublicInbox::GitHTTPBackend::serve($env, $git, $path) : r404();
 }
 
 sub mbox_results {